R440, Astronomy-Mathematics Building, NTU
(台灣大學天文數學館 440室)
Alternating Direction Method of Multipliers and Connections to Split Bregman
Yu-Cheng Zhang (National Central University)
Abstract
The alternating direction method of multipliers (ADMM) is an algorithm that solves convex optimization problems by breaking them into smaller pieces, each of which are then easier to handle. ADMM has attracted renewed attention recently due to its applicability to various machine learning and image processing problem. In this talk, we will introduce this method and the connection between ADMM and the split Bregman algorithm.
Reference
S. Boyd, N. Parikh, E. Chu, B. Peleato, and J. Eckstein, Distributed optimization and statistical learning via the alternating direction method of multipliers, Foundations and Trends in Machine Learning, Volume 3 Issue 1, 2011.